Intestinal dysplasia induced by simian virus 40 T antigen is independent of p53

Abstract

Transgenic mice expressing simian virus 40 large T antigen in enterocytes develop intestinal hyperplasia that progresses to dysplasia with age. Hyperplasia is dependent on T antigen binding to the retinoblastoma (pRb) family of tumor suppressor proteins. Mice expressing a truncated T antigen that inactivates the pRb-family, but is defective for binding p53, exhibit hyperplasia but do not progress to dysplasia. We hypothesized that the inhibition of the pRb family leads to entry of enterocytes into the cell cycle, resulting in hyperplasia, while inactivation of p53 is required for progression to dysplasia. Therefore, we examined T antigen/p53 complexes from the intestines of transgenic mice. We found that T antigen did not induce p53 stabilization, and we could not detect T antigen/p53 complexes in villus enterocytes. In contrast, T antigen expression led to a large increase in the levels of the cyclin-dependent kinase inhibitor p21. Furthermore, mice in which pRb was inactivated by a truncated T antigen in a p53 null background exhibited intestinal hyperplasia but no progression to dysplasia. These data indicate that loss of p53 function does not play a role in T antigen-induced dysplasia in the intestine. Rather, some unknown function of T antigen is essential for progression beyond hyperplasia.

FIG. 1.

Domain map of SV40 T antigen (amino acids 1 to 708) and the mutant dl1137 (amino acids 1 to 121) indicating the three known transformation-related functions of T antigen: the J domain, the retinoblastoma family binding motif, and the p53-binding domain. The phenotypes indicated are from references , , and .

FIG. 2.

Murine intestines expressing T antigen, but not dl1137, exhibit enlarged crypts compared to tissue from nontransgenic littermates. Histological sections from the middle portions of small intestines were prepared and stained with hematoxylin and eosin from control (A, B), TAg^dl1137 (C, D), and TAg^wt (E, F) animals. The left panels show an original magnification of ×200. Enlarged prints of the crypt area are presented on the right panels. Hyperplasia of the TAg^dl1137 sample (C) and dysplasia of the TAg^wt sample (E) are readily observed. TAg^wt and control mice were between 8 and 9 months old, while the TAg^dl1137 mouse depicted in this figure was over 14 months old.

FIG. 3.

TAg^wt have longer intestines than TAg^dl1137 or control mice. (A) The total length of the small intestine was measured and is represented versus the age of the animal. Male (circles) and female (squares) from nontransgenic control (open circles, n = 85; open squares, n = 90), TAg^dl1137 (filled circles, n = 21; filled squares, n = 36) and TAg^wt (gray-shaded circles, n = 33; gray-shaded squares, n = 52) genotypes were analyzed. (B) Increase in intestinal weight of adult TAg^wt (open squares, n = 20) versus control littermates (open squares, n = 16) is evident as the animals age. (C) Ratio between intestinal weight and intestinal length of the small intestine along the life span of adult TAg^wt (gray-shaded squares, n = 20) versus control littermates (open squares, n = 16). (D) Measurements of the intestinal length of nontransgenic and TAg^wt mice were grouped according to age (0 to 3 months, >3 to 6 months, >6 to 9 months, >9 to 12 months, >12 to 15 months, and >15 months). The average intestine length for each group is depicted, the error bars indicating the standard deviation in the group. Corresponding P values and numbers of animals used in each group are indicated.

FIG. 4.

Steady-state levels of p53-pathway-related proteins in TAg^wt mouse intestines or mouse embryo fibroblasts (MEFs). Levels of downstream targets p21 and Mdm2 are also shown. Immunoblot analysis was performed on total soluble protein extracts (30 μg per lane) isolated either from the intestines of 2-month-old nontransgenic or TAg^wt transgenic mice or from control nontransformed or TAg^wt-transformed MEFs. A total of 1 μg of protein extract from Mdm2-overexpressing SF9 cells was loaded as a control for the Mdm2 immunoblot. These results are representative of experiments performed with at least three pairs of mice.

FIG. 5.

Localization of p21 in murine intestines. p21 is restricted to the proliferative and differentiation zones in nontransgenic mice, while p21 expression extends to the upper portions of the villi in murine intestines expressing T antigen. Sections are from 7.6-month-old male mice. These results corroborate the increase of p21 observed in TAg^wt intestines, as shown by immunoblot analysis in Fig. 4. Original magnification was ×200 (×400 for TAg^wt detail).

FIG. 6.

p53 protein is not detected in the intestinal epithelium. (A) Micrograph of one villus with crypts attached to its base (original magnification is ×40), adjacent to a diagram of the same structures as they are separated during the fractionation process. The epithelial regions denoted between the lines indicate approximate regions of enrichment in each fraction: (V) villus stalks and tips; (V/C) intervillus regions, residual villi and crypts; (C) crypts; (M) remaining mesenchyme and muscle. (B) Micrographs of villi and crypts from 2-month-old nontransgenic or TAg^wt mice isolated using the fractionation method described in Materials and Methods, illustrating the enrichment of villus (V) and crypt (C) material. Higher magnification (originally ×100) of the crypt fractions illustrates the increase in crypt length in T antigen-expressing mouse intestines compared to nontransgenic controls. (C) Immunoblot analysis from epithelial fractions isolated from nontransgenic or TAg^wt mouse intestines or from mouse embryo fibroblasts (MEFs) as control. Alkaline phosphatase, lysozyme, and vimentin are markers of villus enterocytes, crypt Paneth cells, and mesenchymal myofibroblasts, respectively, illustrating the enrichment of these markers in their respective fractions. These results are representative of similar fractions from at least four different pairs of mice.

FIG. 7.

SV40 large T antigen is not bound to p53 in the intestinal epithelium of transgenic mice. T antigen was immunoprecipitated from 90 μg of total soluble protein extracts isolated from the small intestinal epithelium of either nontransgenic or TAg^wt transgenic mice or from T antigen-transformed (TAg^wt) MEFs, using a monoclonal antibody PAb419 (25) that specifically recognizes the amino-terminal region of T antigen. Immunoblots were performed on total precipitates, one-sixth of the supernatants, and one-third of the inputs. These results are representative of similar immunoprecipitations from four different pairs of mice or three different clones of TAg^wt MEFs.

FIG. 8.

Analysis of transcriptional levels in fractionated cell populations of the murine small intestine. The middle sections of the small intestine from 3-month-old mice were fractionated and used for RT-PCR as described in Materials and Methods. cDNAs from villus (V), crypt (C), and muscle (M) fractions were amplified using primers for p53, intestinal alkaline phosphatase (Alk. Phos.), l-mannan binding protein (L-MBP), EphB3, SM22α, and GAPDH as described in Materials and Methods. Nonsaturating PCR conditions were used to ensure linearity of the reaction. The products were resolved on 2.5% NuSieve 3:1 agarose gels and stained with GelStar (BioWhittaker Molecular Applications). These results are representative of similar RT-PCRs of fractions from at least four different pairs of mice.